Academic Figure Prompt
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides plain-text links to established design and color palette resources (e.g., adobe.com, coolors.co, colorhunt.co, and colorbrewer2.org) to help users select appropriate academic color schemes. These are well-known technology and design services and do not represent a security risk.
- [INDIRECT_PROMPT_INJECTION]: The skill includes functionality to process and analyze content from external documents provided by users.
- Ingestion points: Instructions in
SKILL.md(Step 1) describe reading user-provided LaTeX, Word, and PDF files to extract research concepts. - Boundary markers: The skill does not explicitly instruct the agent to use delimiters or ignore embedded instructions when processing text from these external documents.
- Capability inventory: The skill's platform configuration (
allowed-tools) includes access tobash, though the provided instructions do not trigger any shell commands. - Sanitization: No explicit sanitization or filtering logic is defined for the document content before it is processed by the model.
Audit Metadata